model.layers.0.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.0.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.0.mlp.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.6250 calibrator=MaxCalibrator quant) model.layers.0.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.0.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.0.mlp.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.6250 calibrator=MaxCalibrator quant) model.layers.0.mlp.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.0.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.0.mlp.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.0.mlp.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.0.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.1.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.1.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.1.mlp.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3125 calibrator=MaxCalibrator quant) model.layers.1.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.1.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.1.mlp.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3125 calibrator=MaxCalibrator quant) model.layers.1.mlp.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.1.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.1.mlp.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.1.mlp.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.1.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.2.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.2.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.2.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.2.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.2.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.2.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.2.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6953 calibrator=MaxCalibrator quant) model.layers.2.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0183 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0176 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0162 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0186 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0187 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0150 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0204 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0183 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0183 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0160 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0172 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0178 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0160 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0187 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0188 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0160 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0197 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0184 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0179 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0208 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0209 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0199 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0188 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0210 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0178 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0189 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0137 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0192 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0212 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0177 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0186 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0192 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0171 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0197 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0199 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0194 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0198 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0182 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0186 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0162 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0200 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0181 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0212 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0209 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0201 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0170 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0176 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0170 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0195 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0205 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0193 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0184 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0160 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0164 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0168 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0136 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0198 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0182 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0184 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0200 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0178 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0192 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0194 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0172 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0168 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0153 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0170 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0176 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0173 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0175 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0189 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0210 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0201 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0189 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0164 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.3125 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.layers.2.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.2.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.3.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.3.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.3.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.3.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.3.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.layers.3.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.layers.3.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.3.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0171 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0171 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0166 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0172 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0167 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0175 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0154 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0157 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0194 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0190 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0197 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0198 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0161 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0188 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0187 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0175 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0167 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0204 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0188 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0182 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0210 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0171 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0177 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0167 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0175 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0155 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0164 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0201 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0167 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0194 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0188 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0170 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0184 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0172 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0157 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0165 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0168 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0166 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0176 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0166 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0154 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0171 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0131 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0171 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0157 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0172 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0160 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0171 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0188 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0199 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0188 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0184 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0170 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0167 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0170 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0159 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0162 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0143 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0166 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0166 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0176 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0186 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0189 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0172 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0154 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0176 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0182 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0199 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0203 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0184 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0182 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0173 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0176 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0203 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0203 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0193 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0187 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0203 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0183 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0181 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0162 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0170 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0168 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0159 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9453 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0172 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0178 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0157 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0177 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0175 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0154 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0167 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0171 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0209 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0168 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0173 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0166 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0177 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0197 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0182 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0162 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0206 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0188 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0190 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0183 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0205 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6758 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0186 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0181 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0203 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0171 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.1250 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0165 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0154 calibrator=MaxCalibrator quant) model.layers.3.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.3.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.4.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.4.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.4.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.4.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.4.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.4.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.layers.4.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.4.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7109 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3887 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.4.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.4.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.5.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.5.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.5.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.5.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.5.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.layers.5.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.layers.5.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6836 calibrator=MaxCalibrator quant) model.layers.5.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.0000 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.5.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.5.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.6.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.6.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.6.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.6.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.6.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.6.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.layers.6.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4375 calibrator=MaxCalibrator quant) model.layers.6.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6992 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0234 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7773 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7617 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6523 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6836 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9688 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7500 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2266 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1172 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7148 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6719 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8594 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.6.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.6.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.7.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.7.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.7.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.7.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.7.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.7.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.7.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2578 calibrator=MaxCalibrator quant) model.layers.7.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9883 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1172 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7188 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7109 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7070 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9922 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0234 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7500 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7891 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3887 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7305 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6719 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4941 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8047 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9062 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.0938 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8984 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9727 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9883 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7773 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9883 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4941 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8047 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7930 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6836 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8438 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4941 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0859 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8477 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.8750 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.7.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.7.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.8.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.8.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.8.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.8.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.8.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.layers.8.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.8.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3516 calibrator=MaxCalibrator quant) model.layers.8.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8164 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0469 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9219 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6719 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7812 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7656 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7188 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9648 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0703 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8594 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7422 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8555 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1562 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0312 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6094 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9609 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1406 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0312 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7305 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0938 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.2500 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.8.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.8.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.9.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.9.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.9.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.9.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.9.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.9.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.layers.9.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8164 calibrator=MaxCalibrator quant) model.layers.9.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1484 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4941 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.8125 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.9.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.9.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.10.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.10.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.10.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.10.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.10.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.10.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.layers.10.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4297 calibrator=MaxCalibrator quant) model.layers.10.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7070 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7656 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0312 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8867 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4941 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9219 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.9922 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6523 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.8047 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8125 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9258 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6523 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8281 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8555 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3750 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.10.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.10.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.11.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.11.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.11.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.11.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.11.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.11.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.11.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9961 calibrator=MaxCalibrator quant) model.layers.11.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9766 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7266 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6992 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9492 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9570 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8008 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7812 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8242 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7969 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8398 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6953 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7812 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6992 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8125 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2656 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8711 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9180 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8008 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8008 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7500 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8555 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8047 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8555 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7930 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7305 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7539 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7188 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9844 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7148 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9922 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7070 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8281 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9883 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9023 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8789 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0469 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8555 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8086 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9805 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9844 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8359 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8594 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6328 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8750 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5625 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8242 calibrator=MaxCalibrator quant) model.layers.11.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.11.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.12.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.12.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.12.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.12.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.12.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.12.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.layers.12.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0469 calibrator=MaxCalibrator quant) model.layers.12.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7969 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8164 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7422 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1250 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8945 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0212 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9297 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6836 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8359 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7969 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7930 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7188 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9375 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7656 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8789 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6953 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7891 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0156 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4766 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6523 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7148 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7656 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8125 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9648 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8867 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6758 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1328 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7109 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8320 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7773 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7891 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0212 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8242 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9648 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9883 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4766 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7070 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9102 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6992 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.6875 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.12.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.12.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.13.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.13.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.13.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.13.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.13.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.13.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.13.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6953 calibrator=MaxCalibrator quant) model.layers.13.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0312 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7969 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4219 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6094 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.8594 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5000 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3594 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1484 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9844 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0078 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8281 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3516 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0938 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4609 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0938 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1797 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0391 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0210 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9531 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8750 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6523 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0209 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3906 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0000 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6523 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1484 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1484 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1250 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0209 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0859 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3516 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9531 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7188 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7891 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1016 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7773 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0212 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9414 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7656 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9219 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8516 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7109 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.2656 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1562 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0703 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9258 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9805 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0938 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1797 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0312 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5000 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1172 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2266 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5078 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0078 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9766 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9883 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7070 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8633 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2734 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4375 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2266 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7422 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0234 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3906 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8633 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1875 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2422 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1484 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8242 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3359 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7383 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7969 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9805 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4609 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5938 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7305 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6562 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0625 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4219 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8477 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9375 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8516 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7031 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2266 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1562 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9844 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1016 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4062 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.4688 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1641 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1250 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8281 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9297 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2266 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7070 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5547 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9922 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2344 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9492 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0938 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9805 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0234 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1172 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6836 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1250 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8828 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.0625 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9141 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9453 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5312 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0312 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9141 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.5312 calibrator=MaxCalibrator quant) model.layers.13.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.13.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.14.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.14.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.14.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.14.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.14.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.layers.14.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.14.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4141 calibrator=MaxCalibrator quant) model.layers.14.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7578 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7188 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8008 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1250 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7656 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7266 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9180 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7344 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0156 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4941 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0234 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1094 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7383 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9492 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7969 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0078 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3594 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7422 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0938 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6758 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2734 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2812 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7070 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7578 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7500 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7930 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7617 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1016 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0469 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8672 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8008 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0210 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6719 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9375 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9883 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9883 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8164 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7305 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8711 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8555 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7578 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0781 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7500 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8477 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9883 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6172 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7500 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0000 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9727 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.2188 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1641 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9180 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6758 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0206 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9219 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8789 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0212 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0234 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0781 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7930 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8242 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9492 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3125 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8984 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6953 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8008 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9961 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9336 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.5625 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.14.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.14.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.15.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.15.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.15.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.15.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.15.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.layers.15.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.layers.15.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.15.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7969 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0212 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7383 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7070 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7422 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0209 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1641 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8047 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7930 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7617 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7148 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6992 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4941 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1094 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8711 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8125 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6836 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8008 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5000 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8594 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4766 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7188 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2109 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6719 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7617 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.2500 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.15.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.15.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.layers.16.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.16.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.16.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.16.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.16.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.layers.16.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.16.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7539 calibrator=MaxCalibrator quant) model.layers.16.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8164 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7812 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3887 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3887 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7539 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7305 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8789 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9180 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0547 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.5000 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.16.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.16.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.layers.17.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.17.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.17.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.17.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.17.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.layers.17.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.layers.17.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.17.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4766 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7383 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8750 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.17.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.17.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.18.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.18.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.18.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.18.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.18.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.18.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.18.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6953 calibrator=MaxCalibrator quant) model.layers.18.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8633 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8438 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.18.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.18.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.19.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.19.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.19.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.19.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.19.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.19.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.19.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.19.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.0312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.19.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.19.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.20.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.20.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.20.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.20.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.20.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.layers.20.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.20.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.20.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.6094 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.20.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.20.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.21.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.21.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.21.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.21.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.21.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.21.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.21.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4766 calibrator=MaxCalibrator quant) model.layers.21.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.7188 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.21.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.21.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.22.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.22.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.22.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.22.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.22.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.22.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.layers.22.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.22.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7969 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0208 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7109 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.0312 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.22.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.22.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.23.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.23.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.23.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.23.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.23.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.layers.23.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.23.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.23.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4766 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.1875 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.23.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.23.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.24.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.24.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.24.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.24.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.24.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.layers.24.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.24.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.24.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6953 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3887 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5469 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.24.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.24.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.25.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.25.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.25.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.25.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.25.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.25.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.25.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.25.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6719 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3887 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3887 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.5312 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.25.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.25.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.layers.26.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.26.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.26.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.26.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.26.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.layers.26.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.26.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0547 calibrator=MaxCalibrator quant) model.layers.26.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6953 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7148 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0078 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8008 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4062 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8438 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0234 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6992 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9141 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7266 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7656 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7070 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7188 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8086 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8789 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0547 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9453 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8086 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6719 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7305 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7266 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8281 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6523 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8633 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8047 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8672 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7539 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6836 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9180 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1484 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8633 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6758 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0078 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0210 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7109 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8516 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7266 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7617 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9531 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7188 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8594 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8438 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0391 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8672 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0625 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0469 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0234 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.8125 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.26.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.26.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.27.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.27.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.27.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.27.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.27.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.layers.27.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.27.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6719 calibrator=MaxCalibrator quant) model.layers.27.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8633 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9453 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8047 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7266 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9727 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4766 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0212 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.8125 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.27.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.27.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.28.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.28.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.28.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.28.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.28.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.28.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.28.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.28.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4941 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4766 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9375 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.28.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.28.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.29.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.29.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.29.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.29.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.29.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.29.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.layers.29.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.29.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7812 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4766 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6758 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6758 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4941 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6523 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.5625 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.29.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.29.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.30.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.30.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.30.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.30.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.30.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.layers.30.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.layers.30.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.30.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3887 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3887 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.8125 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.30.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.30.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.31.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.31.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.31.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.31.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.31.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.31.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.31.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.31.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.0938 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.31.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.31.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.32.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.32.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.32.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.32.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.32.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.32.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.32.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.2812 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.32.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.32.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.33.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.33.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.33.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.33.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.33.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.layers.33.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.layers.33.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.33.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4766 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.6250 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.33.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.33.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.34.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.34.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.34.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.34.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.34.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.layers.34.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.34.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.34.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1484 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.5625 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.34.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.34.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.35.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.35.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.35.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.35.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.35.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.35.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.35.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.35.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7930 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4766 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7383 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6758 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8086 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4941 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9414 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6523 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8633 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7891 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8281 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7148 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.1250 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.35.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.35.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.36.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.36.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.36.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.36.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.36.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.layers.36.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.36.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8398 calibrator=MaxCalibrator quant) model.layers.36.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9531 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0312 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4062 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9336 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8359 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8242 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7578 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7891 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4609 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0312 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8789 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8984 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9180 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4453 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0938 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0391 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1797 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7109 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0312 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1172 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1562 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7148 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9609 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0312 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0625 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2422 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4141 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6523 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0078 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6836 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6953 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9688 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8555 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9062 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2578 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3125 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7383 calibrator=MaxCalibrator quant) model.layers.36.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.36.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.37.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.37.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.37.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.37.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.37.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.layers.37.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.37.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.37.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9062 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7969 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7812 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9102 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9688 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2891 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1250 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9727 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1719 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7109 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3887 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8359 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3887 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9023 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8789 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6836 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7109 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7539 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7500 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8086 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1641 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4941 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7266 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6953 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9102 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8359 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3887 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8047 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7305 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7617 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.3750 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.37.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.37.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.38.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.38.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.38.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.38.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.38.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.38.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.38.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3047 calibrator=MaxCalibrator quant) model.layers.38.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7656 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7070 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8594 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.2188 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9102 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1328 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0547 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8164 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2344 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9492 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1172 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2969 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0703 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9258 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8477 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7969 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9141 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4941 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7188 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1250 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8281 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7422 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7266 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6719 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8750 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6992 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7773 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9258 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6719 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8242 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0625 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7812 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4297 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8516 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1172 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.2500 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.38.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.38.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.39.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.39.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.39.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.39.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.39.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.39.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.layers.39.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3047 calibrator=MaxCalibrator quant) model.layers.39.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8125 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7344 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7617 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6719 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0859 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7109 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7070 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9766 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1719 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8438 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2734 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2969 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9922 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0781 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9531 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8477 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2344 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8906 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5156 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4844 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4766 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2344 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5547 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8086 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9688 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7773 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6953 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0234 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9375 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0000 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5156 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9961 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7305 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7344 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8008 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0078 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8125 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8789 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0938 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7070 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6758 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3906 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6523 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7617 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0625 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1875 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7422 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8906 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3984 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0391 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2969 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1172 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7188 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2734 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8594 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.0156 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8945 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8867 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9258 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7539 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8398 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8398 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3203 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9023 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8086 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9805 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1250 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2344 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1875 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9531 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6992 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9297 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8477 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0234 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7422 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7578 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6719 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0781 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4219 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.2500 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.39.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.39.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.40.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.40.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.40.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.40.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.40.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.40.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.layers.40.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9492 calibrator=MaxCalibrator quant) model.layers.40.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8164 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8320 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1562 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1875 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9883 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3906 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7266 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2344 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6992 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9453 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8477 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0469 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8359 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4941 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7383 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8086 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4062 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9336 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.4375 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2656 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9688 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8320 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1406 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9258 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3203 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9141 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9453 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8008 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8398 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.9844 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9922 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5156 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7305 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9062 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0156 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0625 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9727 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6836 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8867 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9492 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8320 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0625 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.5625 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9492 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0000 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8945 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8672 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0078 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7617 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6992 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9492 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7344 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1641 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2891 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7891 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8008 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4609 calibrator=MaxCalibrator quant) model.layers.40.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.40.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.41.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.41.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.41.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.41.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.41.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.layers.41.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.layers.41.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4688 calibrator=MaxCalibrator quant) model.layers.41.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7539 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7266 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7148 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9531 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9961 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9805 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7266 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3438 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4531 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7930 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1172 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6758 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6836 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0078 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9453 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7344 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4609 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6992 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0078 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0208 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0312 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8008 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7422 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8359 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0781 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0547 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7656 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7188 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9492 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6992 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0391 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9336 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3887 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7344 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0469 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0469 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8047 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7578 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2422 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1016 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8359 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8672 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8008 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8789 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2344 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7188 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3359 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0703 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9414 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9570 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0212 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4062 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.5625 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.41.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.41.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.42.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.42.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.42.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.42.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.42.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.42.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.42.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7812 calibrator=MaxCalibrator quant) model.layers.42.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9531 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8438 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7148 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2891 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3438 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5234 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8945 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1094 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.9141 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8242 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.9531 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1406 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0938 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2734 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9453 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0703 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7578 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3984 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8906 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3281 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1562 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7266 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0000 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8945 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6016 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1641 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6953 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2109 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8398 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1484 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4766 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0234 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9961 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1484 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1094 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4141 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5469 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7773 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8477 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0938 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1406 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7773 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1250 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3594 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3125 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9219 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9062 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5469 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3281 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8438 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5156 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.9453 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0781 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1797 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2031 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9609 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8398 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8242 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2031 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2031 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8164 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3047 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9609 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8281 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0859 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8281 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0859 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9453 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7969 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.0312 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0703 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2656 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2344 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2422 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3594 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2969 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7344 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1172 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0625 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6758 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8359 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.9922 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8281 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1172 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3125 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8633 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.2500 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6719 calibrator=MaxCalibrator quant) model.layers.42.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.42.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.43.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.43.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.43.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.43.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.43.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.43.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.layers.43.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7891 calibrator=MaxCalibrator quant) model.layers.43.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7891 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1406 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2344 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6992 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9453 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1094 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8398 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2266 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8867 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9688 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0210 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2812 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7578 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.0938 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9805 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8438 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7422 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0859 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8047 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0469 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4941 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0078 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4844 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8477 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9922 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8555 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7539 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8633 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0938 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0391 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7305 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9297 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7070 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7344 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8672 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7070 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6992 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5391 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9648 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9062 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0938 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1406 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7422 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9883 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1562 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9180 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4062 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6836 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1797 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0234 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7344 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0938 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9141 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1094 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1562 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3438 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7148 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7773 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9023 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.2500 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8594 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9297 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8516 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8594 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7539 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1172 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9922 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6719 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.43.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.43.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.44.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.44.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.44.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.44.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.44.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.44.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.44.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.4531 calibrator=MaxCalibrator quant) model.layers.44.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7109 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4688 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1797 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7188 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8789 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8789 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6953 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3672 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8789 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3516 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1875 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7578 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2812 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6758 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0212 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9570 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7930 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9336 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1328 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9883 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1250 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1016 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2812 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7656 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6836 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.8047 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5234 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6719 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1250 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8555 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9258 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0000 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0156 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4375 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7539 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8047 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4297 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.2500 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.3125 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7344 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3203 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9141 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4297 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2734 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0703 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9102 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8672 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7344 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9062 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.9062 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.8594 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7812 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9648 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.8125 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8281 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2500 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7930 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7539 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0210 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7656 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1328 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7109 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0781 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3867 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3984 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7266 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3984 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5234 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7773 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6758 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8711 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6016 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8125 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3887 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8828 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1797 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0703 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2656 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0469 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7031 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9922 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3438 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9023 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.8750 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.44.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.44.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.45.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.45.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.45.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.45.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.45.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.45.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.45.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1094 calibrator=MaxCalibrator quant) model.layers.45.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0078 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8906 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8242 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.0781 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8398 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7109 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0859 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0234 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0312 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9648 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0391 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9922 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5469 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2422 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0469 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6836 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1875 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9297 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8555 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6719 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1094 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0938 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1875 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4141 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8945 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2578 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8320 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8320 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8867 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9609 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8789 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8008 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0391 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8359 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9180 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9023 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0625 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7422 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.8516 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6250 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9727 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7383 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6484 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7578 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2812 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8867 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8711 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1875 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0625 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8125 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.8984 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7383 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8867 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9883 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4453 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7383 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1016 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6953 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9766 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8477 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5703 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0859 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8594 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8906 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4219 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6484 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8594 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4844 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.2500 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0234 calibrator=MaxCalibrator quant) model.layers.45.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.45.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.46.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.46.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.46.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.46.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.46.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.46.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.46.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1641 calibrator=MaxCalibrator quant) model.layers.46.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6719 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8750 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8867 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9531 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4219 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7734 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0547 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2188 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7969 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5547 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.1719 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8828 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5625 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1328 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1094 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8750 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6016 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0234 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7383 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.1875 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7500 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1328 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9336 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0312 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8320 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2578 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0212 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8164 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9023 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0547 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5078 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9570 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4844 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7930 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0078 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0156 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4141 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8438 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9844 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1328 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0391 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9414 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4219 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8125 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8477 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4766 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8516 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8789 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8516 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0859 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8398 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9375 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3906 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1562 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1484 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0206 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9141 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7344 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8789 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5312 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7539 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3125 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7148 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1172 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.0000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9609 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.9219 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5781 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2188 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0859 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.5000 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.46.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.46.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.47.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.47.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.47.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.47.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.47.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.47.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.layers.47.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5625 calibrator=MaxCalibrator quant) model.layers.47.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6719 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6523 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5781 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0000 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1641 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8555 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9062 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6758 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0212 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7305 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8164 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3281 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0212 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8359 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8125 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0703 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0212 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9297 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2344 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.0625 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8711 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1016 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8906 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8398 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8086 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1484 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7188 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0205 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7656 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8672 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9922 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1719 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0209 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8945 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3438 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8516 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8633 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7305 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8164 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1406 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6836 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7070 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8281 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8945 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2656 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1172 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4375 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6992 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5234 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0208 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8164 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9961 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4766 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7188 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8945 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7969 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9531 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9688 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7422 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7422 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9844 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.6250 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7383 calibrator=MaxCalibrator quant) model.layers.47.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.47.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.48.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.48.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.48.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.48.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.48.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.48.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.48.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7891 calibrator=MaxCalibrator quant) model.layers.48.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7656 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4941 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0201 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8672 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7891 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7148 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8125 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.9922 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7969 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3203 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8555 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8242 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7578 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7617 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3750 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9844 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1562 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7109 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7109 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9414 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2266 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7891 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7305 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0210 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1641 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2188 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9453 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2109 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.1562 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0078 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8047 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6719 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5938 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8203 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0206 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2734 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8711 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8672 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7969 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8633 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8438 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1406 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.3906 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5234 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7539 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0208 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8633 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3359 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6250 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9062 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3047 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4219 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9648 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2891 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1016 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8633 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.1875 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.48.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.48.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.49.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.49.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.49.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.49.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.49.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.49.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.49.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4141 calibrator=MaxCalibrator quant) model.layers.49.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4844 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2422 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6992 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0234 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7070 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9102 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7109 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9961 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8633 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7344 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7109 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3281 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6953 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0781 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1953 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8242 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6445 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.9141 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1094 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0209 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7656 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8242 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4062 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5156 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9258 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6484 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6992 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8398 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9219 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0210 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0859 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.0938 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9297 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0000 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4844 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8711 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6953 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7891 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2578 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7148 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8281 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8398 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7148 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7539 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1562 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0547 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9297 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7461 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9258 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1172 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4219 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3359 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4688 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7773 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9648 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3750 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7578 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6953 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8281 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2422 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3359 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9688 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.6094 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0212 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1172 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7734 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.9375 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3047 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.6641 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7930 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8867 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9961 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1484 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2031 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3125 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4609 calibrator=MaxCalibrator quant) model.layers.49.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.49.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.50.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.50.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.50.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.50.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.50.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.layers.50.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.50.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8789 calibrator=MaxCalibrator quant) model.layers.50.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0234 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8477 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4141 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3203 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3438 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7305 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6523 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7031 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7891 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3906 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.7500 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2344 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8320 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2031 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7383 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.8125 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9492 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8906 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9297 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9922 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8711 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7617 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8438 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4219 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0391 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8750 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7344 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7930 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3652 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9023 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7656 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7305 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.3359 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1406 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8086 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0469 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9844 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.8281 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8555 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8828 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7344 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.0625 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.50.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.50.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.51.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.51.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.51.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.51.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.51.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.51.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.51.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2422 calibrator=MaxCalibrator quant) model.layers.51.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8477 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9180 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6836 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.4062 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7344 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.7344 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8828 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4297 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6602 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8164 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9961 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6953 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1562 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0938 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0206 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9609 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7930 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0469 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0209 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0781 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9609 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3359 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4941 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7148 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8711 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7891 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0209 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4141 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6523 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7539 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9102 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7500 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7305 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0208 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6289 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=10.3750 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4512 calibrator=MaxCalibrator quant) model.layers.51.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.51.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.52.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.52.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.52.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.52.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.52.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.52.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.layers.52.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2578 calibrator=MaxCalibrator quant) model.layers.52.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6875 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4746 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3516 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0212 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0198 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7539 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4062 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8945 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6523 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0391 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1562 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.5938 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7578 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4102 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8984 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8789 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9141 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0212 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8477 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9219 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0208 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7852 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8086 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7773 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0703 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0209 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2578 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6953 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9922 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2344 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7656 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4980 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8359 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6914 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6484 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9531 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6172 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3887 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1953 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.1719 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7344 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.9219 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7227 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7695 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4766 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7266 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8516 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4473 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0156 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0206 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.9062 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7188 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7383 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.0781 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4551 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0398 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.5312 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5742 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7617 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4766 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6133 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6641 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0216 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.1562 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7070 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6680 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.2109 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.7188 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4785 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0232 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4453 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.9688 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.4297 calibrator=MaxCalibrator quant) model.layers.52.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.52.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.53.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.53.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.53.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.53.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.53.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.53.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.53.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.53.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4824 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4922 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8906 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5938 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0210 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0206 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0439 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5430 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6094 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5703 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6250 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4766 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5508 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4238 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6211 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4883 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5469 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3887 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5078 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5000 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4902 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5898 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5859 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0223 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4043 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5547 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6562 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6367 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4609 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0311 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6055 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0205 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3262 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5625 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4121 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0210 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4277 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5586 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0217 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3164 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6797 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3984 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3535 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6328 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5820 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6016 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.4375 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.53.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.53.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.layers.54.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.54.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.54.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.54.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.54.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.54.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.layers.54.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.layers.54.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4160 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0210 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0330 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0211 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0221 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0215 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0432 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0288 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0400 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0437 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0210 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0210 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0214 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0295 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0422 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0381 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5391 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0226 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0420 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0220 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4844 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0234 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0256 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3438 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0444 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0361 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0228 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0325 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0264 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4336 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4375 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0227 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0322 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0237 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=12.8750 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.layers.54.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.54.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.layers.55.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.q_proj.weight_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.k_proj.weight_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.v_proj.weight_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.o_proj.weight_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.gate_proj.input_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.gate_proj.weight_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.layers.55.self_attn.softmax_quantizer TensorQuantizer(disabled) model.layers.55.mlp.router.gate.input_quantizer TensorQuantizer(disabled) model.layers.55.mlp.router.gate.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.router.gate.weight_quantizer TensorQuantizer(disabled) model.layers.55.mlp.shared_experts.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.shared_experts.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.shared_experts.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.55.mlp.shared_experts.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.shared_experts.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.shared_experts.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.55.mlp.shared_experts.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.55.mlp.shared_experts.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.shared_experts.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5039 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.0.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.0.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.0.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.0.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.0.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.0.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.0.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.0.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.0.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.1.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.1.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.1.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.1.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.1.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.1.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.1.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.1.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.1.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.2.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.2.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.2.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.2.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.2.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.2.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.2.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0197 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.2.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.2.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.3.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.3.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.3.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.3.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.3.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.3.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.3.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.3.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.3.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4668 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.4.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.4.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.4.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.4.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.4.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.4.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0306 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.4.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0251 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.4.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.4.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.5.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.5.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.5.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0261 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.5.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.5.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.5.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.5.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0204 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.5.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.5.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.6.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.6.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.6.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.6.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.6.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.6.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0229 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.6.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0262 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.6.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.6.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.7.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.7.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.7.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.7.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.7.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.7.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.7.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.7.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.7.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.8.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.8.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.8.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.8.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.8.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.8.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.8.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.8.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.8.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.9.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.9.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.9.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.9.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.9.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.9.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.9.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.9.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.9.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.10.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.10.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.10.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.10.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.10.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.10.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0430 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.10.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.10.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.10.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.11.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.11.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.11.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.11.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.11.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.11.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.11.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.11.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.11.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.12.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.12.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.12.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.12.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.12.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.12.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0222 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.12.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.12.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.12.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.13.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.13.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.13.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0250 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.13.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.13.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.13.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.13.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.13.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.13.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.14.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.14.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.14.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0378 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.14.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.14.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.14.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.14.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.14.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.14.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.15.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.15.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.15.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.15.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.15.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.15.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0249 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.15.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0391 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.15.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.15.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.16.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.16.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.16.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.16.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.16.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.16.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0248 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.16.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.16.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.16.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.17.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.17.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.17.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.17.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.17.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.17.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.17.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.17.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.17.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.18.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.18.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.18.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.18.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.18.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.18.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.18.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.18.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.18.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.19.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.19.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.19.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.19.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.19.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.19.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.19.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0344 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.19.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.19.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.20.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.20.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.20.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.20.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.20.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.20.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.20.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0374 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.20.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.20.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.21.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.21.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.21.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.21.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.21.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.21.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.21.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.21.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.21.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.22.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.22.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.22.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.22.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.22.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.22.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.22.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.22.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.22.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.23.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.23.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.23.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.23.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.23.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.23.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.23.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.23.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.23.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.24.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.24.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.24.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.24.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.24.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.24.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.24.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.24.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.24.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.25.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.25.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.25.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.25.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.25.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.25.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.25.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.25.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.25.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.26.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.26.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.26.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.26.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.26.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.26.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.26.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.26.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.26.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.27.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.27.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.27.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.27.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.27.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.27.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.27.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.27.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.27.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.28.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.28.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.28.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.28.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.28.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.28.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.28.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.28.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.28.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.29.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.29.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.29.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.29.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.29.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.29.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.29.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.29.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.29.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.30.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.30.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.30.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.30.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.30.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.30.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0225 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.30.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.30.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.30.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.31.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.31.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.31.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0305 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.31.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.31.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.31.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0236 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.31.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.31.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.31.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.32.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.32.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.32.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.32.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.32.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.32.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.32.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.32.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.32.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.33.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.33.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.33.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.33.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.33.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.33.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.33.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0198 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.33.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.33.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.34.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.34.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.34.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.34.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.34.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.34.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.34.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.34.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.34.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.35.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.35.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.35.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.35.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.35.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.35.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.35.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.35.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.35.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.36.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.36.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.36.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.36.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.36.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.36.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.36.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0427 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.36.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.36.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.37.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.37.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.37.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.37.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.37.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.37.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.37.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.37.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.37.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.38.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.38.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.38.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.38.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.38.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.38.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.38.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.38.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.38.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.39.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.39.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.39.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.39.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.39.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.39.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.39.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.39.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.39.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.40.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.40.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.40.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.40.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.40.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.40.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0260 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.40.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.40.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.40.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.41.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.41.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.41.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.41.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.41.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.41.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.41.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0413 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.41.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.41.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.42.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.42.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.42.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.42.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.42.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.42.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.42.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.42.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.42.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.43.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.43.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.43.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.43.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.43.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.43.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0298 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.43.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.43.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.43.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.44.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.44.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.44.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.44.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.44.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.44.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.44.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.44.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.44.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.45.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.45.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.45.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.45.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.45.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.45.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.45.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0277 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.45.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.45.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.46.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.46.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.46.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.46.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.46.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.46.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.46.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.46.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.46.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3320 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.47.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.47.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.47.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.47.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.47.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.47.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0435 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.47.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.47.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.47.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.48.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.48.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.48.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.48.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.48.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.48.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.48.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.48.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.48.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.49.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.49.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.49.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.49.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.49.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.49.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.49.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.49.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.49.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.50.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.50.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.50.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.50.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.50.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.50.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.50.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.50.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.50.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.51.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.51.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.51.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.51.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.51.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.51.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.51.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.51.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.51.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.52.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.52.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.52.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0310 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.52.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.52.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.52.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0243 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.52.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.52.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.52.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.53.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.53.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.53.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0286 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.53.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.53.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.53.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.53.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.53.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.53.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.54.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.54.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.54.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.54.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.54.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.54.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.54.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0292 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.54.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.54.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.55.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.55.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.55.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.55.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.55.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.55.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0339 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.55.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.55.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.55.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.56.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.56.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.56.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0386 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.56.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.56.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.56.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.56.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.56.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.56.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.57.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.57.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.57.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.57.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.57.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.57.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.57.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.57.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.57.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.58.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.58.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.58.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.58.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.58.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.58.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.58.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.58.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.58.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.59.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.59.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.59.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.59.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.59.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.59.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.59.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0317 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.59.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.59.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.60.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.60.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.60.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.60.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.60.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.60.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.60.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.60.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.60.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.61.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.61.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.61.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.61.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.61.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.61.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.61.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0334 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.61.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.61.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.62.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.62.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.62.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.62.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.62.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.62.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.62.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.62.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.62.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.63.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.63.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.63.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0452 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.63.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.63.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.63.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.63.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.63.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.63.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.64.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.64.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.64.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0354 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.64.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.64.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.64.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.64.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0233 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.64.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.64.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.65.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.65.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.65.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0255 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.65.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.65.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.65.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0275 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.65.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0208 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.65.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.65.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.66.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.66.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.66.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.66.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.66.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.66.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.66.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.66.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.66.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.67.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.67.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.67.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.67.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.67.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.67.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.67.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.67.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.67.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.68.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.68.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.68.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.68.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.68.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.68.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0327 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.68.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0273 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.68.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.68.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.69.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.69.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.69.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0244 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.69.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.69.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.69.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.69.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0281 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.69.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.69.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.70.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.70.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.70.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.70.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.70.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.70.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.70.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.70.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.70.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.71.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.71.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.71.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0272 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.71.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.71.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.71.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0238 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.71.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.71.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.71.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.72.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.72.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.72.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.72.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.72.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.72.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.72.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0302 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.72.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.72.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.73.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.73.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.73.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0352 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.73.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.73.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.73.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.73.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0303 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.73.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.73.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.74.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.74.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.74.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.74.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.74.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.74.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.74.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.74.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.74.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.75.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.75.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.75.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.75.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.75.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.75.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.75.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.75.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.75.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.76.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.76.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.76.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0405 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.76.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.76.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.76.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0270 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.76.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.76.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.76.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.77.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.77.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.77.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.77.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.77.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.77.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.77.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.77.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.77.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.78.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.78.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.78.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0294 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.78.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.78.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.78.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.78.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.78.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.78.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.79.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.79.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.79.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.79.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.79.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.79.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.79.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.79.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.79.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.80.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.80.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.80.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.80.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.80.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.80.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0356 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.80.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.80.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.80.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.81.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.81.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.81.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.81.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.81.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.81.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0240 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.81.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.81.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.81.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.82.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.82.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.82.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0258 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.82.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.82.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.82.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0239 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.82.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.82.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.82.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.83.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.83.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.83.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.83.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.83.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.83.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.83.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.83.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.83.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.84.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.84.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.84.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.84.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.84.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.84.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.84.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.84.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.84.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.85.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.85.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.85.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.85.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.85.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.85.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.85.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.85.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.85.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.86.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.86.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.86.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.86.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.86.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.86.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.86.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.86.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.86.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.87.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.87.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.87.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.87.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.87.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.87.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.87.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0376 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.87.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.87.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.88.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.88.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.88.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0282 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.88.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.88.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.88.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0291 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.88.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0242 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.88.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.88.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.89.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.89.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.89.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.89.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.89.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.89.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.89.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.89.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.89.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.90.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.90.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.90.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.90.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.90.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.90.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.90.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.90.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.90.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.91.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.91.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.91.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.91.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.91.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.91.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.91.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.91.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.91.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.92.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.92.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.92.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.92.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.92.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.92.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.92.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0337 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.92.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.92.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.93.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.93.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.93.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.93.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.93.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.93.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.93.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.93.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.93.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.94.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.94.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.94.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0278 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.94.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.94.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.94.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.94.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.94.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.94.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.95.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.95.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.95.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0383 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.95.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.95.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.95.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.95.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.95.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.95.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.96.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.96.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.96.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.96.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.96.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.96.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.96.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0447 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.96.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.96.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.97.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.97.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.97.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.97.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.97.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.97.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.97.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.97.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.97.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.98.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.98.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.98.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0267 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.98.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.98.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.98.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0259 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.98.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0410 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.98.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.98.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.99.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.99.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.99.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0396 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.99.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.99.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.99.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.99.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.99.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.99.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.100.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.100.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.100.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.100.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.100.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.100.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.100.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.100.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.100.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.101.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.101.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.101.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.101.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.101.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.101.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0312 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.101.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.101.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.101.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.102.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.102.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.102.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0293 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.102.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.102.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.102.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0245 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.102.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0308 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.102.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.102.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.103.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.103.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.103.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.103.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.103.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.103.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.103.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.103.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.103.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.104.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.104.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.104.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.104.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.104.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.104.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.104.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.104.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.104.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.105.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.105.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.105.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0371 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.105.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.105.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.105.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0309 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.105.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.105.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.105.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.106.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.106.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.106.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.106.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.106.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.106.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.106.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0284 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.106.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.106.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.107.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.107.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.107.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.107.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.107.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.107.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.107.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.107.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.107.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.108.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.108.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.108.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.108.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.108.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.108.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.108.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.108.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.108.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.109.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.109.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.109.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.109.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.109.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.109.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.109.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.109.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.109.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.110.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.110.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.110.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0297 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.110.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.110.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.110.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0269 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.110.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0266 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.110.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.110.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.111.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.111.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.111.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0315 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.111.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.111.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.111.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0231 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.111.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0332 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.111.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.111.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.112.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.112.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.112.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0364 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.112.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.112.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.112.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0369 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.112.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.112.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.112.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.113.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.113.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.113.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.113.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.113.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.113.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0347 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.113.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.113.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.113.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.114.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.114.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.114.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0254 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.114.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.114.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.114.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0320 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.114.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.114.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.114.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.115.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.115.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.115.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.115.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.115.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.115.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0449 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.115.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0289 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.115.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.115.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.116.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.116.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.116.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.116.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.116.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.116.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.116.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0425 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.116.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.116.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.117.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.117.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.117.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0299 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.117.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.117.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.117.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0408 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.117.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0388 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.117.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.117.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.118.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.118.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.118.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.118.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.118.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.118.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0247 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.118.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0300 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.118.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.118.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.119.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.119.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.119.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.119.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.119.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.119.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.119.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.119.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.119.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.120.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.120.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.120.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0359 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.120.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.120.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.120.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0287 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.120.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.120.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.120.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.121.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.121.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.121.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.121.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.121.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.121.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0219 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.121.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0415 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.121.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.121.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.122.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.122.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.122.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0366 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.122.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.122.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.122.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0393 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.122.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.122.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.122.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.123.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.123.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.123.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0280 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.123.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.123.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.123.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0253 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.123.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0265 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.123.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.123.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.124.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.124.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.124.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.124.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.124.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.124.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0271 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.124.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0349 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.124.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.124.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.125.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.125.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.125.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0304 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.125.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.125.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.125.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0403 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.125.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0283 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.125.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.125.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.126.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.126.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.126.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.126.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.126.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.126.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0342 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.126.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.126.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.126.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.127.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.127.gate_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.127.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0276 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.127.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.5625 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.127.up_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.127.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.127.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0417 calibrator=MaxCalibrator quant) model.layers.55.mlp.experts.127.down_proj.output_quantizer TensorQuantizer(disabled) model.layers.55.mlp.experts.127.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) lm_head.input_quantizer TensorQuantizer(disabled) lm_head.output_quantizer TensorQuantizer(disabled) lm_head.weight_quantizer TensorQuantizer(disabled) 63941 TensorQuantizers found in model